Skip to content

A thought on the hot hand in basketball and the relevance of defense

I was reading about basketball the other day and a thought came to me about the hot hand . . . There are a bunch of NBA players who could shoot with great accuracy even from long distance if they’re not guarded, right? For example, what would Steph Curry’s success rate be for 30-footers if he weren’t guarded and had time to take a shot? Over 50%?

So then it seems like one way to get a hot hand is to not be guarded, or, more generally, to not be guarded tightly. And I guess this would be even more the case for close-in shots, which just about any NBA player could make at a near-100% rate if there were no defense.

This all suggests that a key part—maybe the most important part—of the hot hand is what defense is on you, and more generally how you handle the defense.

I’m not quite sure what more to say about this right now, but this all seems different from the usual way we talk about the hot hand, with a focus on shooting. It also suggests that it could be a mistake to consider free-throw shooting as somehow a more pure test of the hot hand. If the above speculations are correct, then the hot hand in free-throw shooting is really a completely different thing than the hot-hand for regular shots.


  1. Stefan says:

    Slightly related is a paper that tries to study this defense mechanism (ie automatically balancing the defense to good players, so that all of them have an equal success rate). Could be an interesting read (even though I personally have some concerns):

  2. Jonathan (another one) says:

    Sure, but all shooting is volitional. If you’re covered, no one one makes you shoot. The hot hand covers not just your shooting capabilities, but your shooting judgment as well. So if I miss a shot because I decide to take one when I was better defended, then I wasn’t really hot before, just open and competent.

    • Jonathan (another one) says:

      Taking this logic another step…. No one would argue that a 30% jump shooter who happened to get 10 straight backdoor layups had turned “hot.” At least they wouldn’t argue this watching it, although it would look in the data (unless some adjustment is made for degree of difficulty) that the given player was on fire.

    • Z says:

      Well they do call it “hot hand” and not “hot judgment” so I think interest lies more in the physical action once the decision is made, controlling for the degree of difficulty of the shot. There are metrics in basketball statistics meant to reflect shot difficulty that can be adjusted for, e.g. how close the nearest defender is, or just binary challenged vs not challenged. These aren’t perfect of course.

  3. jd says:

    Just a general comment on the hot hand, as I feel that most athletes assume the hot hand is real, no matter their sport. As a former athlete who competed at a decently high level (road bicycle racing), I can say that there are some days that, for unknown reasons, you perform better or worse than ‘average’ (no news here, this happens in work, school, etc). Just in case readers don’t realize though, unlike the rest of life, many elite athletes treat themselves as an experiment of n=1, all the time. They try to do the same thing, with no variation, before each performance – same amount of sleep, same food, same taper routine in the workouts in the weeks prior, same fluids, same warmup, same music, etc. etc. etc., so much so that this can lead to the appearance of superstition. Even though such efforts are made to run the same ‘experiment’ each time (and thereby reduce variance in performance), performance varies, and it can do so even over the same race course or opponents. There are likely a plethora of small physiological reasons for this, and small things make a large difference on performances that hinge on extremely small errors. Even in endurance sports, a minuscule change can make a big difference. In cycling, lets say that the eventual winner of a 5% 5km mountain finish at the end of a bike race is 65kg with a 7kg bike. He attacks and sustains 20km/h up the climb. This is ~388w on average (using Bike Calculator). If the same size person averages 1% less watts, that translates to ~8sec at the finish, which looks like a large margin. If you have ever looked at a power meter while riding a bike, 4watts is such a small difference that you can’t even keep it within the error of your pedal stroke (i.e. it’s not possible to hold the power at a given number + or – 4 watts). Could this not also be the “hot hand” manifested differently? I imagine that in a sport such as basketball that puts a premium on neuromuscular hand-eye coordination, there could also be a plethora of small physiological factors that could add up into the appearance of a hot hand or a slump, and these might change more rapidly since they involve more CNS factors than say endurance performance on a bike. Is it not plausible that these changes could occur minute by minute or hour by hour if it is something that involves more of a neuromuscular precision like shooting a basket?

    Apologies for the excessively long post, but it does seem like one of those cases where there are so many variables, most of them unknown, as to be difficult to prove or disprove with the data that is available (at least in the sense of understanding it), as the data that is available doesn’t seem likely to contain the relevant variables.

  4. Javier says:

    The hothand is clearly bullshit and you should be worried by any arguments that lead you to conclude it isn’t.

  5. Terry says:

    This sort of thing comes up in economics all the time. Other parties adjust to the actions of one party, so the observed profit or whatever is a complex function of all the actions and reactions.

    Generally, though you usually don’t get the result that all advantages or profits are neutralized or driven to zero, and in this case, I would not expect shooting success to be equalized over all players (speaking loosely here).

    This is because there is a cost to the defense to guard the hot player more closely. The optimum defense, therefore is probably for the defense to put some extra effort into defending the hot player more closely, but not enough to completely equalize shooting success across all players.

    In fact, this seems almost certainly to be the case if we think beyond the hot hand and think more generally about how some players are just better than others all the time. Yes, you might guard Michael Jordan more closely because he is a better player, and this will attenuate Jordan’s performance, but it isn’t done enough to equalize performance across all the players on Jordan’s team. We just don’t see that.

Leave a Reply